Summary: A Two-Round Variant of EM for Gaussian Mixtures
Sanjoy Dasgupta

AT&T Labs ­ Research
Leonard J. Schulman
Georgia Institute of Technology
Abstract
We show that, given data from a mixture of
¡
well-separated spherical Gaussians in ¢¤£ , a sim-
ple two-round variant of EM will, with high
probability, learn the centers of the Gaussians to
near-optimal precision, if the dimension is high
(¥§¦©¨
¡
). We relate this to previous theoreti-
cal and empirical work on the EM algorithm.
1 Introduction
At present EM is the method of choice for learning mix-
tures of Gaussians. A series of theoretical and experimen-

Source: Abu-Mostafa, Yaser S. - Department of Mechanical Engineering & Computer Science Department, California Institute of Technology

Collections: Computer Technologies and Information Sciences